Authoritarianism, Prejudice, and Support for Welfare Chauvinism in the United States

1 Authoritarianism, Immigration, and Welfare Chauvinism

Welfare policies are ethnically and racially neutral on the policy surface, but welfare programs become a mechanism for “defining structures of social and political inclusion and exclusion” (Lieberman 2003, p. 30) that is reflected in public opinion. The inclusion and exclusion of groups requires the construction of imagined communities creating group boundaries that can be defined in terms of acquiring and holding specific social norms (Anderson 1983). An imagined community creates a sense of bonds and connections to group members, individuals sharing the same social norms, and a willingness to support the wellbeing of in-group members (Miller 1995). For in-group members, inclusive welfare policies provide support, but for out-group members, welfare programs can be viewed as providing resources to create parallel societies allowing them to resist social integration resulting in the continuation of the normative threat presented by outgroups. In this way, support for welfare is predicted by group identities with preferences for limiting access to out-group members by adopting restrictive policies (Kinder and Kam 2009; Stenner 2005; Theiss-Morse 2009).

The terms authoritarian and authoritarianism can be problematic due to the different uses. For example, many individuals picture leaders of certain political regimes when they think about the concepts of authoritarians and authoritarianism, such as Saddam Hussein, Bashar al-Assad, and Vladimir Putin (Hetherington and Weiler 2009). Another use of the terms is to study the differences in human behaviour and psychology (i.e. Adorno et al. 1950; Altemeyer 1981, 2003; Feldman 2003; Hetherington and Weiler 2009; Stenner 2005). For example, Adorno et al. (1950) researched authoritarianism as a personality trait. Their research was motivated to explain the mass support for the antidemocratic policies of Nazi Germany, ethnocentrism, and prejudice. Many of the earlier studies focus on a single dimension of authoritarians by viewing authoritarians as individuals who submit to authority (Altemeyer 1981). Authoritarianism has at least two additional dimensions beyond submission, social conformity and aggression (Altemeyer 1996). More recent studies have focused more on the dimensions of social conformity and aggression (Feldman 2003; Hetherington and Suhay 2011; Hetherington and Weiler 2009; Stenner 2005), and define authoritarianism as a general orientation “concerned with the appropriate balance between group authority and uniformity, on the one hand, and individual autonomy and diversity, on the other hand” (Stenner 2005, p. 14). Under this conceptualization, social conformity, defined as “a desire for an orderly and structured world” (Hurwitz and Peffley 1992, p. 400), is a key motivation for authoritarians’ attitudes and policy preferences. A need for social order among authoritarians leads to a black-and-white view of the world that distinguishes sharply between favoured in-groups and threatening out-groups (Altemeyer 1981 , 1996; Feldman and Stenner 1997; Feldman 2003; Hetherington and Weiler 2009; Katz 1960; Stenner 2005; Walker 2009). Authoritarians are individuals with these three dimensions: a greater desire for social conformity, hold prejudicial attitudes towards “targeted” groups resulting in support for aggressive and punitive policies towards those groups, and support for stronger authority to enforce social conformity to create order (Altemeyer 2003; Hetherington and Weiler 2009).

Media framing shapes attitudes and policy preferences by shaping the considerations people use to form attitudes and the weight they place on those considerations (e.g. Chong and Druckman 2007; Nelson, Clawson, and Oxley 1997). These media influenced considerations are the situtuationalism in how immigration is viewed by the masses, which policies people connect to immigration, and the type of threat that natives perceive from immigrants (Chavez 2008; Lahav and Courtmanche 2012; Lahav 2013). The news media can create a false perception that immigrants are “unwilling or incapable of integrating, of becoming part of the national community” (Chavez 2008, p. 2) resulting in “the failure of Mexican-Americans over several generations to increase their identification with American values” (Huntington 2004, p. 242) being a commonly held belief. As a result, immigrants that fail to adopt social norms can be viewed as a threat to the present social order and as being outside of the “American” imagined community (Paxton and Mughan 2006; Schildkraut 2011). When a frame triggers a pre-existing and often used predisposing factor, such as authoritarianism, individuals are more likely to use that predisposing factor to interpret and react to the issue. For example, there is considerable evidence that the public views immigration as a racialized^[2] issue and, as such, whites often conflate their views of Latinos and immigrants, and draw on their racial stereotypes to form immigration attitudes (e.g. Brader, Valentino and Suhay 2008; Burns and Gimpel 2000; Chavez 2008; Lu and Nicholson-Crotty 2010).

In this line of thought, authoritarianism acts as a worldview and a fundamental consideration that links strong emotional responses, such as fear and anxiety, about out-groups to government policies resulting in enduring political blocs. Authoritarians are much more likely than non-authoritarians to express prejudice towards a range of outgroups that are perceived to challenge the existing social and political order, such as immigrants. Perceptions of social differences between in-groups and out-groups provide the fertile breeding ground for authoritarians’ negative views of out-groups and a preference for largely punitive policies that restrict the social, economic and political rights of outgroups to minimize perceived threats. For immigrants and Latinos, group boundaries help to “rationalize and justify governmental practices and policies that stigmatize and punish certain categories of immigrants and their children” (Chavez 2008, p. 42). These preferred policies include decreasing immigration levels and restricting government benefits and civil liberties of groups perceived to be racially or culturally different from “true” Americans (Adorno et al. 1950; Feldman 2003; Hetherington and Weiler 2009; Hetherington and Suhay 2011; Kinder and Kam 2009; Stenner 2005). As a result, authoritarians would have a preference for welfare policy “to become more aggressive in enforcing social order, defining behavioural expectations, and demanding self-discipline from the poor” (Soss, Fording and Schram 2011, p. 63). Based on this research, I hypothesize that immigrant groups that are branded as alien out-groups will be targeted by authoritarians for a series of harsh and punitive measures, such as creating waiting periods before immigrants can access welfare programs.

Authoritarianism and racial stereotypes are not mutually exclusive but entangled influences of political attitudes. As such, authoritarianism can be expressed through negative stereotypes forming prejudice towards out-groups. Stereotypes are conditional “pictures in the head” of labels representing perceived group characteristics that are used for relevant judgements (Hurwitz and Peffley 1997; Lippmann 1922). Racial stereotypes influence political attitudes when the policy relates to perceptions about a particular minority groups (Nelson and Kinder 1996). For example, Burns and Gimpel argue that “The association of immigrants with welfare use and the connection between welfare use and negative racial stereotypes underscore how immigration can easily be translated into a racial issue with redistributive undertones” (2000, p. 207). This body of literature mirrors the studies on welfare preferences finding that racial stereotypes and concerns for social conformity undermine support for generous welfare programs (Gilens 1999; Hopkins 2009; Hurwitz and Peffley 1992; Johnson 2001; Peffley, Hurwitz, and Sniderman 1997). Previous research finds that authoritarians hold more negative stereotypes of several minority groups than non-authoritarians, such as Jews and African Americans (e.g. Sniderman and Piazza 1993; Hetherington and Wieler 2009; Stenner 2005). On the other hand, authoritarianism has been shown to be a weak predictor of a racial stereotype measure of ethnocentrism (Kinder and Kam 2009). Theoretically, authoritarians are probably concerned about the conformity of immigrants beyond standard measures of racial stereotypes, such as being hard working. The basis for authoritarian support of restrictive immigration policies can be due to the nature of immigrants being social and cultural outsiders, not just prejudicial stereotypes of them. In this possible scenario, authoritarians’ preferences should remain the same regardless of their prejudice of Latinos.

2 Data, Measurement, and Analysis

The data for the analysis are from the ANES in 1992. The 1992 ANES includes several important features. First, I am unaware of another American survey that includes a measure of authoritarianism and welfare chauvinism. Second, the ANES uses the child-rearing questions to measure authoritarianism (Hetherington and Suhay 2011; Hetherington and Weiler 2009; Stenner 2005). The child-rearing measure has several advantages, such as being politically neutral, over other measures of authoritarianism. Unfortunately, the 1992 ANES has several disadvantages. First, ANES dataset lacks an adequate number of blacks and Latinos. Due to this restriction, I limit the samples to non-Latino whites. Second, the data is old but for several reasons, I do believe that the worldview divide found in this study would be significantly stronger in recent years. First, the worldview divide between authoritarianism and non-authoritarianism is believed to be constant or “timeless” (Hetherington and Weiler 2015). For example, research on authoritarianism and terrorism by Hetherington and Suhay (2011) find that authoritarians’ attitudes and policy preferences towards terrorist suspects do not vary with the level of perceived threat. Second, additional research finds that authoritarianism is a constant predictor of many policy preferences across time (Cizmar et al. 2014). Third, over the last few decades the polarization in American politics has increased indicating that the divide on welfare chauvinism today is most likely to be greater than it was in 1992. Authoritarians have increasingly sorted into the Republican Party (Hetherington and Weiler 2009; Cizmar et al. 2014). This sorting has resulted in a growing divide between the Democratic and Republican parties over immigration, and as such, the divide I find in this paper would probably be larger in 2020 as non-authoritarian Democrats adopt more positive immigration attitudes. These findings support the theory that authoritarianism is a stable worldview that should have a strongly consistent influence on public opinion.

3 Results

In the first section of the analysis, I primarily focus on the question, are authoritarians more likely to support welfare chauvinism than non-authoritarians? My results provide evidence supporting a hypothesis that a gap in preferences for welfare chauvinism exists between authoritarians and non-authoritarians. Authoritarians are more supportive of a waiting period with 85.2% holding this specific policy preference, while only 73.1% of non-authoritarians desire a waiting period (significant at p-value <0.001).^[9] The authoritarian gap is about 12.1%. The correlation between authoritarianism and welfare chauvinism suggests a relationship with a value of 0.14 (p-value <0.001). In sum, this provides basic evidence that authoritarianism does shape preferences for welfare chauvinism.

In the next step of the analysis, I investigate if authoritarianism continues to predict support for welfare chauvinism after adding control variables because support for welfare chauvinism can be shaped by confounding factors that potentially influence the effect of authoritarianism. Authoritarianism, as a predisposition or worldview, holds an early and important position in the formation of political attitudes (Feldman 2003; Hetherington and Weiler 2009; Stenner 2005). As a result, authoritarianism may have an indirect effect on welfare chauvinism through other attitudes. Table 1 presents a series of models starting with authoritarianism and basic demographics (Model 1A). From this baseline model, I add “downstream” variables that may mediate the influence of authoritarianism. In Model 1B, I include political ideology and partisanship to the baseline. The third column, Model 1C, displays the model with addition of the prejudice measure. For my analysis and discussion of these results, I primarily focus on the influence of authoritarianism and prejudice on support for immigrant welfare chauvinism.

Starting with the first three models, authoritarianism is positively and significantly related at the 0.001 level to welfare chauvinism. This pattern of coefficients for authoritarianism is consistent with my expectations, that authoritarians are more likely to support a waiting period before immigrants can access welfare programs. The substantive impact of authoritarianism on welfare chauvinism can be estimated using CLARIFY to simulate predicted probabilities (King, Tomz, and Wittenberg 2000; Tomz, Wittenberg, and King 2003). For each model, I estimate the first differences for authoritarianism while holding all other variables constant.^[10] The predicted probabilities for Models 1A and 1B are roughly the same. For the baseline model, Model 1A, authoritarians are 17.3% more likely to support an immigrant waiting period. Controlling for ideology and partisanship, in Model 1B, results in basically the same predicted probability, 17.2%, with the political variables held at mean values. In sum, the addition of political identifications does not alter the relationship between authoritarianism and welfare chauvinism. Authoritarianism is not operating through these political variables, at least not in 1992.

Consistent with previous research, I find that authoritarians are more likely to hold negative stereotypes of minorities (e.g. Hetherington and Weiler 2009; Stenner 2005). In this case, authoritarians are more likely to hold prejudicial views of Latinos (r = −0.19, p < 0.001) providing support that authoritarians are more likely to be prejudiced towards Latinos than non-authoritarians. Based on these basic results, authoritarians may express their dislike for an out-group through prejudice. To further explore this relationship, I include the Latino prejudice measure, which is negatively and significantly related to welfare chauvinism (p < 0.001), in Model 1C. Individuals who hold a negative perception of Latinos’ work-ethic, relative to stereotypes of whites, are more likely to prefer exclusive welfare policies for immigrants. The authoritarianism coefficient remains significant at the 0.01 level. The substantive effect of authoritarianism on welfare chauvinism remains the same with a predicted probability of 17.4% for a first difference change in authoritarianism while holding all other variables constant and prejudice at its mean value. The predicted probability for authoritarianism does decrease to 9.4% among the most prejudiced individuals in the sample. The authoritarian gap is the largest between authoritarians and non-authoritarians who more positive stereotypes of Latinos with authoritarians being 24.2% more likely to support a waiting period before immigrants can access welfare programs.

These results indicate that the influence of authoritarianism is greatest among those with the least prejudice and that prejudicial authoritarians and non-authoritarians hold more similar policy preferences.

4 Discussion and Conclusion

Several findings from my study have implications. First, the scholarly literature has long studied the importance of political and social attitudes, such as ideology, prejudice, economics, and core values, as predictors of individual support for immigration and welfare preferences (e.g. Gilens 1999; Hainmueller and Hiscox 2007; Peffley, Hurwitz, and Sniderman 1997). Additional studies link authoritarianism to immigration and welfare preferences (e.g. Hetherington and Weiler 2009; Rickert 1998). My results support the extension of many of these studies from welfare preferences in general to preferences for a waiting period before immigrants can receive benefits. In sum, these preferences go beyond political considerations, economic considerations, and prejudice but these attitudes are shaped by considerations regarding social conformity.

Second, the scholarly literature has long studied the importance of prejudice in the formation of political attitudes finding that prejudice can be extremely powerful in shaping policy preferences towards redistributive policies (e.g. Fording, Soss, and Schram 2011; Fox 2004; Gilens 1999; Johnson 2001; Peffley, Hurwitz, and Sniderman 1997) and immigrants (Brader, Valentino and Suhay 2008; Burns and Gimpel 2000; Chavez 2008; Lu and Nicholson-Crotty 2010). In a similar area of research, scholars find that authoritarianism is more likely to hold prejudicial stereotypes (Adorno et al. 1950; Feldman 2003; Hetherington and Weiler 2009; Stenner 2005). My findings indicate that prejudice has the largest impact on the authoritarian gap among the control variables.^[11] The largest authoritarian gap exists among individuals with the least prejudicial attitudes of Latinos. Non-authoritarians that view Latinos negatively are only slightly less likely than authoritarians to support a waiting period requirement for access to welfare benefits. In sum, authoritarianism has its largest effect among the least prejudiced individuals.

Third, a significant percentage, 73.2%, of non-Latino whites with a non-authoritarian worldview prefers creating a waiting period before immigrants can access welfare programs. In the early 1990s, political elites and media frames connected immigration as a threat to domestic spending programs, including welfare (Calavita 1996). Given the disproportionate number of media stories that picture Latinos as the face of contemporary immigration and portray immigrants as failing to adopt American social norms (Chavez 2008; Haynes, Merolla, and Ramakrishnan 2016) a preference for restricting access to welfare can gain support among non-authoritarians. While my findings are limited in scope due to data restricting the analysis to a single time point, the research on issue salience suggests that non-authoritarians may adopt restrictive immigration preferences during periods when immigration is portrayed by the media and political elites as a normative threat. During these salient periods of time, authoritarians and threatened non-authoritarians may form a political base that can potentially undermine liberal immigration policies through an opinion-policy linkage.